Reported is a study to determine if cloze tests are sensitive to gain measurement using two different approaches to science instruction. A secondary purpose was to determine what relationships existed among the scores of 300 fifth-grade students (150 boys, 150 girls), randomly selected from a population of 520 fifth-grade students in Warminster, Pennsylvania. The instructional program covered three weeks and included 160 minutes per week for both groups. Control group teachers taught traditionally using a basic science textbook. Experimental group teachers used individualized science activities developed by a Bethlehem Area School District ESEA Title III program, Keys for Interdisciplinary Tutorial Study. A two factor analysis of variance was used to determine differences between gain scores made by pupils of both groups. The relationships among general reading ability, science ability and gain scores on cloze tests were determined by the Pearson Product Moment coefficient of correlation. Conclusions drawn included: (1) the cloze test instruments were sensitive to gain measurement; and (2) individualized approach to science instruction was not superior to the traditional science textbook approach in terms of cloze test gain scores. (Author/EB)